Evolve or Die: How I Stopped Avoiding and Starting Loving Python Upgrades
summary
A tool called version up scans lockfiles and requirements.txt, flags deprecated stdlib modules and package imports, and estimates time to upgrade; demo shows NumPy still had ~2M downloads on 3.7 and 1.5M on 3.8 despite deprecation, and warns 3.8 deprecates in months. Data sits in Snowflake as parsed stdlib and package tables so teams can run SQL to find which packages (e.g., setuptools uses imp) and which third‑party deps introduce risk, enabling concrete upgrade tasks and exported Jira cards.
AI-written summary · about
chapters (6)
- 0:00Python upgrades urgency
- 1:54Delaying upgrades costs
- 3:44Barriers to upgrading Python
- 8:00Risk aversion in upgrades
- 12:59Tools for upgrade certainty
- 24:44Key takeaways
related talks
32mHow (not) to upgrade a platform
34mKeeping up with Python: what makes upgrades hard, and what can we do about it
26mPython 2 to 3: How to Upgrade and What Features to Start Using
35mUpgrading Legacy Projects: Lessons Learned
27mHow We Migrated 3.8 Million Lines of Python 2 Without Interrupting Deve...
26mPython 2 is dead! Drag your old code into the modern age
30mThe Evolutionary Saga of Python Packaging
45mWriting Python 2/3 compatible code